Streaming Average Score per Landing Page

An ads analytics pipeline processes a stream of post-click quality rating events and computes the running average quality score per landing page. Each event contains a user ID, landing page name, advertiser vertical, and a quality score. Process events one at a time (streaming) and compute the average quality score per landing page across all events seen so far. **Input:** `events` — a list of `(user_id, page_name, vertical, score)` tuples in arrival order. **Output:** A dict mapping each landing page name to its average quality score (float) across all events.